Table 4.

5-fold cross-validation results for recurrence prediction using different ensemble learning models.

Model	Acc	Recall	Prec	F1 score
LightGBM	0.7600	0.7673	0.7733	0.7553
	± 0.0579	± 0.1311	± 0.0771	± 0.0716
CatBoost	0.6833	0.6164	0.7107	0.6511
	± 0.0543	± 0.1313	± 0.0397	± 0.0879
XGBoost	0.7224	0.6346	0.8032	0.6936
	± 0.0834	± 0.1154	± 0.1521	± 0.0978
GBDT	0.6543	0.6382	0.6600	0.6387
	± 0.0463	± 0.1328	± 0.0286	± 0.0828

Each model was evaluated employing the mean of each fold result and corresponding 95% CI.